def attempt_dispatch(expt_config, expt_dir, chooser, driver, options):
'''
Dispatches a job containing `num_jobs` jobs, if the number of jobs is greater
than 1 they will all have the same proc_id.
'''
log("\n" + "-" * 40)
if isinstance(expt_config, str):
expt = load_experiment(expt_config)
else:
expt = expt_config
# Build the experiment grid.
expt_grid = ExperimentGrid(expt_dir,
expt.variable,
options.grid_size,
options.grid_seed)
jobs = []
num_jobs = 1
for n in range(num_jobs):
# Print out the current best function value.
best_val, best_job = expt_grid.get_best()
if best_job >= 0:
log("Current best: %f (job %d)" % (best_val, best_job))
else:
log("Current best: No results returned yet.")
# Gets you everything - NaN for unknown values & durations.
grid, values, durations = expt_grid.get_grid()
# Returns lists of indices.
candidates = expt_grid.get_candidates()
pending = expt_grid.get_pending()
complete = expt_grid.get_complete()
n_candidates = candidates.shape[0]
n_pending = pending.shape[0]
n_complete = complete.shape[0]
log("%d candidates %d pending %d complete" %
(n_candidates, n_pending, n_complete))
# Verify that pending jobs are actually running, and add them back to the
# revisit this.
# candidate set if they have crashed or gotten lost.
#for job_id in pending:
# proc_id = expt_grid.get_proc_id(job_id)
# if proc_id != -1 and not driver.is_proc_alive(job_id, proc_id):
# log("Set job %d back to candidate status." % (job_id))
# expt_grid.set_candidate(job_id)
# Track the time series of optimization.
write_trace(expt_dir, best_val, best_job, n_candidates, n_pending, n_complete)
# Print out the best job results
write_best_job(expt_dir, best_val, best_job, expt_grid)
if n_complete >= options.max_finished_jobs:
log("Maximum number of finished jobs (%d) reached."
"Exiting" % options.max_finished_jobs)
return 0, None
if n_candidates == 0:
log("There are no candidates left. Exiting.")
return 0, None
#Don't launch unless we can launch the complete bundle.
if n_pending >= options.max_concurrent or (n == 0 and n_pending + num_jobs > options.max_concurrent):
log("Maximum number of jobs (%d) pending." % (options.max_concurrent))
return 1, None
else:
# Ask the chooser to pick the next candidate
log("Choosing next candidate... ")
time_cand_start = time.time()
job_id = chooser.next(grid, values, durations, candidates, pending, complete)
time_cand = time.time() - time_cand_start
log("Chose a candidate (took %i secs)." % (time_cand))
# If the job_id is a tuple, then the chooser picked a new job.
# We have to add this to our grid
if isinstance(job_id, tuple):
(job_id, candidate) = job_id
job_id = expt_grid.add_to_grid(candidate)
log("selected job %d from the grid." % (job_id))
# Convert this back into an interpretable job and add metadata.
job = Job()
job.id = job_id
job.expt_dir = expt_dir
job.name = expt.name
job.language = expt.language
job.status = 'submitted'
job.submit_t = int(time.time())
job.param.extend(expt_grid.get_params(job_id))
if options.nb_mini_batches > 0:
batch_i = expt_grid.mini_batch_i
expt_grid.mini_batch_i = (batch_i + 1) % options.nb_mini_batches
batch_param = Parameter()
batch_param.name = 'batch_i'
batch_param.int_val.append(batch_i)
job.param.extend([batch_param])
save_job(job)
if num_jobs == 1:
pid = driver.submit_job(job)
if pid != None:
log("submitted - pid = %s" % (pid))
expt_grid.set_submitted(job_id, pid)
else:
log("Failed to submit job!")
log("Deleting job file.")
os.unlink(job_file_for(job))
else:
jobs.append(job)
#Temporary, we don't have a proc id yet.
expt_grid.set_submitted(job_id, -1)
#Delayed submit when there is more than one job bundled.
if num_jobs > 1:
pid = driver.submit_job(jobs)
if pid != None:
log("Submitted %i jobs with pid = %s" % (num_jobs, pid))
for j in jobs:
expt_grid.set_submitted(j.id, pid)
else:
log("Failed to submit job!")
log("Deleting job files.")
for j in jobs:
os.unlink(job_file_for(j))
return 2, pid
def attempt_dispatch(expt_config, expt_dir, chooser, driver, options):
log("\n" + "-" * 40)
expt = load_experiment(expt_config)
# Build the experiment grid.
expt_grid = ExperimentGrid(expt_dir,
expt.variable,
options.grid_size,
options.grid_seed)
# Print out the current best function value.
best_val, best_job = expt_grid.get_best()
if best_job >= 0:
log("Current best: %f (job %d)" % (best_val, best_job))
else:
log("Current best: No results returned yet.")
# Gets you everything - NaN for unknown values & durations.
grid, values, durations = expt_grid.get_grid()
# Returns lists of indices.
candidates = expt_grid.get_candidates()
pending = expt_grid.get_pending()
complete = expt_grid.get_complete()
n_candidates = candidates.shape[0]
n_pending = pending.shape[0]
n_complete = complete.shape[0]
log("%d candidates %d pending %d complete" %
(n_candidates, n_pending, n_complete))
# Verify that pending jobs are actually running, and add them back to the
# candidate set if they have crashed or gotten lost.
for job_id in pending:
proc_id = expt_grid.get_proc_id(job_id)
if not driver.is_proc_alive(job_id, proc_id):
log("Set job %d back to pending status." % (job_id))
expt_grid.set_candidate(job_id)
# Track the time series of optimization.
write_trace(expt_dir, best_val, best_job, n_candidates, n_pending, n_complete)
# Print out the best job results
write_best_job(expt_dir, best_val, best_job, expt_grid)
if n_complete >= options.max_finished_jobs:
log("Maximum number of finished jobs (%d) reached."
"Exiting" % options.max_finished_jobs)
return False
if n_candidates == 0:
log("There are no candidates left. Exiting.")
return False
if n_pending >= options.max_concurrent:
log("Maximum number of jobs (%d) pending." % (options.max_concurrent))
return True
else:
# start a bunch of candidate jobs if possible
#to_start = min(options.max_concurrent - n_pending, n_candidates)
#log("Trying to start %d jobs" % (to_start))
#for i in xrange(to_start):
# Ask the chooser to pick the next candidate
log("Choosing next candidate... ")
job_id = chooser.next(grid, values, durations, candidates, pending, complete)
# If the job_id is a tuple, then the chooser picked a new job.
# We have to add this to our grid
if isinstance(job_id, tuple):
(job_id, candidate) = job_id
job_id = expt_grid.add_to_grid(candidate)
log("selected job %d from the grid." % (job_id))
# Convert this back into an interpretable job and add metadata.
job = Job()
job.id = job_id
job.expt_dir = expt_dir
job.name = expt.name
job.language = expt.language
job.status = 'submitted'
job.submit_t = int(time.time())
job.param.extend(expt_grid.get_params(job_id))
save_job(job)
pid = driver.submit_job(job)
if pid != None:
log("submitted - pid = %d" % (pid))
expt_grid.set_submitted(job_id, pid)
else:
log("Failed to submit job!")
log("Deleting job file.")
os.unlink(job_file_for(job))
return True
def attempt_dispatch(expt_config, expt_dir, chooser, driver, options):
log("\n" + "-" * 40)
expt = load_experiment(expt_config)
# Build the experiment grid.
expt_grid = ExperimentGrid(expt_dir, expt.variable, options.grid_size,
options.grid_seed)
# Print out the current best function value.
best_val, best_job = expt_grid.get_best()
if best_job >= 0:
log("Current best: %f (job %d)" % (best_val, best_job))
else:
log("Current best: No results returned yet.")
# Gets you everything - NaN for unknown values & durations.
grid, values, durations = expt_grid.get_grid()
# Returns lists of indices.
candidates = expt_grid.get_candidates()
pending = expt_grid.get_pending()
complete = expt_grid.get_complete()
n_candidates = candidates.shape[0]
n_pending = pending.shape[0]
n_complete = complete.shape[0]
log("%d candidates %d pending %d complete" %
(n_candidates, n_pending, n_complete))
# Verify that pending jobs are actually running, and add them back to the
# candidate set if they have crashed or gotten lost.
for job_id in pending:
proc_id = expt_grid.get_proc_id(job_id)
if not driver.is_proc_alive(job_id, proc_id):
log("Set job %d back to pending status." % (job_id))
expt_grid.set_candidate(job_id)
# Track the time series of optimization.
write_trace(expt_dir, best_val, best_job, n_candidates, n_pending,
n_complete)
# Print out the best job results
write_best_job(expt_dir, best_val, best_job, expt_grid)
if n_complete >= options.max_finished_jobs:
log("Maximum number of finished jobs (%d) reached."
"Exiting" % options.max_finished_jobs)
return False
if n_candidates == 0:
log("There are no candidates left. Exiting.")
return False
if n_pending >= options.max_concurrent:
log("Maximum number of jobs (%d) pending." % (options.max_concurrent))
return True
else:
# start a bunch of candidate jobs if possible
#to_start = min(options.max_concurrent - n_pending, n_candidates)
#log("Trying to start %d jobs" % (to_start))
#for i in xrange(to_start):
# Ask the chooser to pick the next candidate
log("Choosing next candidate... ")
job_id = chooser.next(grid, values, durations, candidates, pending,
complete)
# If the job_id is a tuple, then the chooser picked a new job.
# We have to add this to our grid
if isinstance(job_id, tuple):
(job_id, candidate) = job_id
job_id = expt_grid.add_to_grid(candidate)
log("selected job %d from the grid." % (job_id))
# Convert this back into an interpretable job and add metadata.
job = Job()
job.id = job_id
job.expt_dir = expt_dir
job.name = expt.name
job.language = expt.language
job.status = 'submitted'
job.submit_t = int(time.time())
job.param.extend(expt_grid.get_params(job_id))
save_job(job)
pid = driver.submit_job(job)
if pid != None:
log("submitted - pid = %d" % (pid))
expt_grid.set_submitted(job_id, pid)
else:
log("Failed to submit job!")
log("Deleting job file.")
os.unlink(job_file_for(job))
return True
def explore_space_of_candidates(experiment, objective_function,
working_directory, chooser,
grid_size=1000,
grid_seed=1,
max_finished_jobs=100):
# Build the experiment grid.
expt_grid = ExperimentGrid(working_directory,
experiment.variables, grid_size, grid_seed)
next_jobid = 0
run_python_job = PythonRunner()
while next_jobid < max_finished_jobs:
best_val, best_job = expt_grid.get_best()
# Gets you everything - NaN for unknown values & durations.
grid, values, durations = expt_grid.get_grid()
# Returns lists of indices.
candidates = expt_grid.get_candidates()
pending = expt_grid.get_pending()
complete = expt_grid.get_complete()
n_candidates = candidates.shape[0]
n_pending = pending.shape[0]
n_complete = complete.shape[0]
logging.info("%d candidates %d pending %d complete", n_candidates,
n_pending, n_complete)
if n_candidates == 0:
logging.info("There are no candidates left. Exiting.")
return
# Ask the chooser to pick the next candidate
logging.info("Choosing next candidate... ")
job_id = chooser.next(grid, values, durations, candidates, pending, complete)
yield best_val, best_job, expt_grid.get_params(best_job), job_id
# If the job_id is a tuple, then the chooser picked a new job.
# We have to add this to our grid
if isinstance(job_id, tuple):
(job_id, candidate) = job_id
job_id = expt_grid.add_to_grid(candidate)
logging.info("selected job %d from the grid", job_id)
expt_grid.set_submitted(job_id, next_jobid)
expt_grid.set_running(job_id)
start_t = time.time()
result, memoized = run_python_job(job_id, objective_function,
expt_grid.get_params(job_id), working_directory)
duration = time.time() - start_t
expt_grid.set_complete(job_id, result, duration)
next_jobid += 1
if memoized:
max_finished_jobs += 1
